Young New Zealanders’ Beliefs About Youth Suicide and How It Can Be Prevented

Mikayla S. Holmana, BSc (Hons), & Matt N. Williamsa, BBS, GradDipArts, MA, PhD

aSchool of Psychology, Massey University, New Zealand

Young New Zealanders’ Beliefs About Youth Suicide and How It Can Be Prevented

Abstract

Objective: New Zealand has the highest suicide rate amongst youth (ages 15-24) in the

OECD. In this study, we aimed to conduct a conceptual replication of two previous studies

(Heled & Read, 2005; Curtis, 2010), examining the views that youth in New Zealand hold about the causes of youth suicide, potential solutions, and help-seeking.

Method: A detailed data collection and analysis plan was preregistered prior to data collection. One hundred university students aged 18 to 24 completed a mixed-methods online survey; 89% were female.

Results: Just one of four hypotheses formulated based on the findings of Curtis (2010) was supported: Students who were personally aware of another student's suicidality were more willing to seek help for others from the university counselling service. Qualitative findings indicated that bullying and stigma were the most commonly perceived causes of youth suicide. Improvement of mental health services was the most frequently recommended solution for reducing the youth suicide rate.

Conclusions: The views of youth should be included in the future development of mental health services and policies aimed at reducing suicide rates for this population.

Keywords: help-seeking, mental health services, suicide, university, youth

ii

1 Suicide is a leading cause of death in many countries, with an estimated 800,000

2 suicide deaths worldwide each year (World Health Organization, 2018). In 2012, New

3 Zealand had the highest suicide rate out of all OECD countries for youth aged 15 to 24

4 (Ministry of Social Development, 2016). Currently, a young person in New Zealand dies by

5 suicide every sixty-four hours. In the year to June 2018, the total number of suicide deaths for

6 10 to 24 year olds was 137, the highest number for the country since 2013 (Coronial Services

7 of New Zealand, 2018).

8 Youth aged 15 to 24 consistently report the highest rates of psychological distress

9 compared to other age groups in New Zealand, with 11.8% of youth scoring 12 or higher on

10 the Kessler Psychological Distress Scale-10 in a Ministry of Health (2017) survey. However,

11 those towards the upper end of the youth age range, the 20 to 24 year olds, are the ones who

12 more often die by suicide (Beautrais, 2003). University students aged under 24 are a

13 particularly vulnerable population, reporting higher levels of psychological distress than non-

14 university populations of the same age (Stallman, 2010). Counselling services offered by

15 universities New Zealand wide are experiencing surges in the number of youths trying to

16 access these services, with an overall increase of nearly 25% between 2015 and 2017 (New

17 Zealand Union of Students' Associations, 2018). These circumstances suggest that suicide

18 prevention services or programmes specifically designed for this population need to be

19 developed.

20 Why Research Young People's Beliefs About Youth Suicide?

21 The Mental Health Commission (2012) reports that service users and their families

22 frequently express the need for inclusion in the development of policies and services. This

23 inclusion promotes self-determination, empowering individuals to share the decision making

24 process with professionals, and increasing treatment adherence (Corrigan et al., 2012). Self-

25 determination is a crucial factor in whether service use will result in recovery (Mental Health

BELIEFS ABOUT YOUTH SUICIDE

26 Commission, 2012), and has implications for suicide prevention specifically. Understanding

27 the beliefs young people hold about suicide could impact mental health services targeted

28 towards this population, and in combination with the knowledge of mental health

29 professionals, the content and delivery of such services could be adapted to increase their

30 relevance and effectiveness for youth. This has the potential to save the lives of more young

31 New Zealanders who access these services during times of distress.

32 At present there are a small number of studies in New Zealand that have attempted to

33 research beliefs about youth suicide among university students. Two of these studies that are

34 of particular relevance to this study are by Heled and Read (2005) and Curtis (2010).

35 Heled and Read (2005) researched youth perceptions of suicide and its solutions.

36 Undergraduate students at the University of Auckland were asked for possible reasons for

37 why the youth suicide rate was high and suggestions for how the rate could be reduced, with

38 students providing qualitative responses to these questions. Participants believed the greatest

39 cause of youth suicide to be pressure from adults or peers, followed by financial worries and

40 poor job prospects. A tenth of the participants perceived that insufficient, poorly advertised,

41 and inaccessible support services contributed to the high rate of youth suicide. Participants

42 suggested that to reduce the suicide rate, public awareness was important. Other suggestions

43 included creating support groups and increasing the availability of and access to counselling.

44 Curtis (2010) examined perceptions of youth at Wellington’s Victoria University

45 towards suicide and help-seeking behaviour. Students in this study largely agreed that they

46 would contact a university mental health service for themselves (49.8% strongly agree or

47 agree) and small but significant gender differences were found, with females more likely to

48 seek help for themselves than males were. Students who had their own experience of

49 suicidality were more confident that they could identify an at-risk student, and less likely to

50 believe suicide could be prevented. Groups of students were interviewed in a second phase of

BELIEFS ABOUT YOUTH SUICIDE

51 Curtis' (2010) study, during which many students said that when concerned about another

52 student they would prefer to be self-reliant or seek help from family and friends, rather than

53 turning to professional services such as those offered by the university.

54 Research Aims

55 In light of the findings reported by the two studies above, this study aimed to

56 investigate perceptions of the high youth suicide rate in New Zealand among young

57 university students at Massey University.

58 This study is a conceptual replication of parts of the two studies described earlier,

59 neither of which were preregistered. The strategy of preregistration has grown in popularity

60 in psychology in recent years in response to well-known problems with the reproducibility of

61 psychological studies (see Nosek, Ebersole, DeHaven, & Mellor, 2018; Open Science

62 Collaboration, 2015; Pashler & Wagenmakers, 2012). Preregistering data collection and

63 analysis plans prior to collecting data limits the capacity of the researcher to exploit

64 “researcher degrees of freedom” (Simmons, Nelson, & Simonsohn, 2011, p. 1359) in order to

65 produce statistically significant findings—a practice also known as “p-hacking” (Head,

66 Holman, Lanfear, Kahn, & Jennions, 2015). As such, this study was preregistered on the

67 Open Science Framework (see Method for details), with hypotheses, method and analyses

68 decided upon prior to data collection. The focus of this study was on the perceived causes of

69 youth suicide, potential ways that youth believed could decrease this rate, and their current

70 help-seeking behaviour towards Massey University's counselling service.

71 Both Heled and Read's (2005) and Curtis' (2010) studies were original research and

72 were largely exploratory. Because of this, this study had only a few hypotheses, all of which

73 were based on a small selection of key statistically significant differences found by Curtis.

74 The hypotheses were: 1) female students will be more likely than male students to say they

75 would seek help if experiencing personal, health or work-related issues; 2) students who are

BELIEFS ABOUT YOUTH SUICIDE

76 personally aware of another student's suicidality will be more willing to seek help for others

77 from the university counselling service; 3) students who have been suicidal themselves will

78 feel more confident that they can identify students at risk of suicide; and 4) students who

79 have been suicidal themselves will be less likely to believe that suicide can be prevented.

80 Method

81 Participants and procedure

82 A sample of students from Massey University were recruited for this study. A

83 university population was used in both Heled and Read's (2005) and Curtis' (2010) studies,

84 and it was a convenient, accessible population for this study as well. A questionnaire was

85 used for data collection, which was distributed to students via their undergraduate course

86 websites and social media. As specified in the study's preregistration, accessible at

87 https://osf.io/t3ynv/, the questionnaire remained online for six weeks, during which time 140

88 students responded. Of these responses, there were 100 that met the inclusion requirements of

89 the preregistration (e.g., answered at least half of the questions that were going to be

90 analysed). Only these responses were included in the analyses. The preregistration specified

91 that if there were missing values on quantitative variables, single imputation would be

92 executed. Of the 100 responses included in the analyses, none required imputation for

93 missing values. The overall sample size for the study was only moderate, but delivered

94 adequate power for the correlational analyses that were used to test the majority of

95 hypotheses: An N of 100 provides 86% power to detect a medium-size correlation of  = 0.3

96 in a 2-tailed test.

97 Participant demographics. The age of participants ranged from 18 to 24, with a

98 mean age of 21.38 (SD = 2.00). Females were over-represented in this study, making up 89%

99 of the sample (see Table 1).

100 Measures

BELIEFS ABOUT YOUTH SUICIDE

101 A questionnaire containing 27 items was used for the study. This can be viewed at

102 https://osf.io/n5zc8/. The first five items were related to demographic information, asking

103 participants about their gender, age, ethnicity, what region they live in, and their area of study

104 at Massey University.

105 The following four items were replicated from Heled and Read's (2005) study. The

106 first two of these items asked if participants knew a young person who had attempted suicide,

107 or who had died by suicide. The next two items replicated from Heled and Read asked

108 participants what they perceived the causes of New Zealand's high youth suicide rate to be,

109 and what they thought could be done to reduce the rate (with qualitative responses).

110 Fourteen items were based on items from Curtis' (2010) study, with small changes

111 made to the wording of some items. Participants were asked to rate their level of agreement

112 with statements about their help-seeking behaviour, such as "I would personally seek help

113 from Massey University's Health & Counselling Centre if I was experiencing personal, health

114 or work-related issues". These items were rated on a Likert scale from 1 (strongly agree) to 5

115 (strongly disagree).

116 A further three items were created specifically for this study. The first of these items

117 asked participants what they would do to help a distressed or suicidal friend (with qualitative

118 responses). Another item asked whether participants knew about the services offered by

119 Massey University's Health and Counselling Centre. The last item asked participants to

120 choose what their preferred method of booking an appointment at the Health and Counselling

121 Centre would be: walk-in to make an appointment, phone call, email, Massey University

122 website, or Massey University app.

123 Data sharing policy

124 A de-identified copy of the quantitative data has been transferred to an Open Science

125 Framework project, along with analysis code that was used in R (R Core Team, 2018). This

BELIEFS ABOUT YOUTH SUICIDE

126 can be accessed at https://osf.io/nw8db/. Qualitative responses will not be made openly

127 accessible. It is impossible to guarantee that individual participants would be unidentifiable

128 based on the details in these responses.

129 Ethics approval

130 Approval for this study was given by the Massey University Human Ethics

131 Committee.

132 Results

133 Quantitative results

134 Descriptive statistics. Most participants in this sample personally knew a young

135 person that had attempted suicide (82%), while just over half of all participants personally

136 knew a young person that had died by suicide (55%).

137 The distribution of responses to each Likert item is shown in Figure 1. The Likert

138 scale ranged from 1 (strongly agree) to 5 (strongly disagree), the same scale used by Curtis

139 (2010) for these items. Of the 100 participants in this sample, 29 had personally provided

140 support to students at Massey University that were engaging in suicidal behaviour (M = 3.52,

141 SD = 1.45), and 30 had encouraged fellow students to seek support for suicidal behaviour (M

142 = 3.37, SD = 1.57).

143 For further information about the quantitative results, please see the Supplementary

144 Materials at https://osf.io/khbvg/.

BELIEFS ABOUT YOUTH SUICIDE

145

146 Figure 1. Bar plot of Likert items showing the percentage of each response category.

147 Confirmatory analyses. A Welch’s t test was conducted to test hypothesis 1, that

148 female students would show higher levels of agreement with the statement “I would

149 personally seek help from Massey University’s Health and Counselling Centre if I was

150 experiencing personal, health or work-related issues”. Female participants indicated slightly

BELIEFS ABOUT YOUTH SUICIDE

151 more agreement with this statement (M = 2.70) than males did (M = 2.48). However, this

152 difference was not statistically significant, t(11.19) = 0.52, p = .614, d = 0.17. The robustness

153 of this result was checked with a Mann-Whitney U test (as per preregistration, section C)

154 which also showed a non-significant result, W = 493, p = .570.

155 Pearson's correlation tests were conducted to test hypotheses 2, 3, and 4, followed by

156 Spearman's rank correlation tests to check robustness (as per preregistration, section C).

157 These analyses found some support for hypothesis 2. Being aware of another student's

158 suicidality was positively correlated with being more likely to seek help for peers, Pearson’s r

159 = .255, p = .010; Spearman’s rs = .286, p = .004. The third hypothesis was that students who

160 had themselves been suicidal would be more confident that they could identify a student at

161 risk of suicide. The analyses did not support this hypothesis, with the Pearson and

162 Spearman’s correlation tests both suggesting very small relationships that were not

163 statistically significant, r = -.015, p = .884; rs = .029, p = .771. Hypothesis 4 was also not

164 supported by the results of the correlation tests. Students who had themselves been suicidal

165 were more likely to believe suicide was preventable, not less likely (as was hypothesised),

166 with the analyses finding a weak positive correlation between the two variables, r = .198, p =

167 .048; rs = .210, p = .036.

168 Qualitative results

169 The responses to each of the three qualitative questions were coded one question at a

170 time, using a qualitative content analysis method as described by Forman and Damschroder

171 (2007). This involved reading all responses to the question, followed by highlighting key

172 words or phrases. These excerpts were typed into an Excel document, then all were read

173 again before being coded inductively, with codes broadly describing the content of the

174 excerpt. For example, "lack of conversation around the topic" was highlighted as a key phrase

175 in one participant's response to the question "What do you think are causes of New Zealand's

BELIEFS ABOUT YOUTH SUICIDE

176 high youth suicide rate?" This was later grouped with similar excerpts from other participants

177 like "lack of discussion amongst families, friends and society" and "I don't believe it is talked

178 about enough" to create the broad code "not talking about it". The original responses were

179 then read again to count the number of participants who mentioned this code, crossing off

180 each associated excerpt to ensure all were counted. This process of coding responses occurred

181 for each of the three qualitative questions. The total for each code is reported as the

182 percentage of the sample that mentioned it. As each participant’s response could be assigned

183 to multiple codes, the codes were not mutually exclusive.

184 Participants' responses to the question "What do you think are causes of New

185 Zealand's high youth suicide rate?" largely focused on factors that led to young people feeling

186 suicidal. Bullying was the most commonly stated cause, being mentioned by 32% of

187 participants. Stigma and New Zealand's culture were also thought to be largely at fault,

188 mentioned by around a quarter of participants (23% and 25% respectively). A number of

189 participants perceived poor mental health services to be a cause of the high youth suicide rate,

190 with 14% mentioning that these services were difficult to access, with long wait times and

191 eligibility criteria that were too strict.

192 Participants' responses to the question "What do you think could be done to reduce

193 New Zealand's high youth suicide rate" saw many propose ways to improve mental health

194 services for this age group. 37% of the sample said that providing a different type of service

195 would help to reduce the rate, while 20% thought that increasing accessibility would result in

196 an improvement. The need for education about mental health, suicide, and symptom

197 recognition was also frequently mentioned.

198 In response to the question "If one of your friends was experiencing distress, suicidal

199 thoughts, or suicidal behaviour, how would you go about helping them?" the majority of

200 participants mentioned that they would seek professional help, while just under half would

BELIEFS ABOUT YOUTH SUICIDE

201 offer help, support, advice or guidance. Talking and listening were also frequently perceived

202 to be ways to help a suicidal friend.

203 For further information about the qualitative results, please see the Supplementary

204 Materials at https://osf.io/khbvg/.

205 Discussion

206 Of the four hypotheses, only one was found to be supported. Similar to the findings of

207 Curtis (2010), students who were aware of a peer's suicidality were significantly more likely

208 to seek help for others.

209 There were multiple differences between the present study's results and the findings

210 of Curtis (2010), with the remaining three hypotheses not supported. Curtis found a

211 significant gender difference in how likely students were to seek help for themselves, while

212 the present study did not find this result. This could be due to differences in the demographics

213 of this sample compared to Curtis'. This study had only 10 male participants (10%), limiting

214 the statistical power of this analysis, while males made up 35.2% of Curtis' sample. The third

215 hypothesis was that students who had their own suicidal experience would be more confident

216 in identifying a student at risk of suicide (as reported by Curtis, 2010). This hypothesis was

217 not supported. The fourth hypothesis, that students who had themselves been suicidal would

218 be less likely to believe that suicide could be prevented, was also not supported. Instead the

219 opposite result to Curtis (2010) was found, with these students being significantly more likely

220 to believe suicide could be prevented.

221 Nearly a third of the students in this sample (30%) said they would encourage a

222 student to get support for suicidal behaviour. This proportion was nearly double that found by

223 Curtis (2010), suggesting that students at Massey University may be more open to seeking

224 help than Curtis' sample of Victoria University students were eight years ago. This is possibly

225 due to the large proportion of psychology students in the present study. However, as the

BELIEFS ABOUT YOUTH SUICIDE

226 degree or study major of students in Curtis' (2010) study was not reported, a direct

227 comparison cannot be made.

228 The results of this study show that students at Massey University know of the services

229 offered by the university's Health and Counselling Centre and many would be willing to

230 utilise these services when in need themselves. Despite this willingness to seek help from the

231 university's service, students held quite negative perceptions toward mental health services in

232 general, mentioning barriers such as being inadequate or difficult to access, and having long

233 wait lists or strict entry criteria, with such views repeated in other literature (e.g., Heled and

234 Read, 2005).

235 In order to rectify some of these barriers to seeking help from mental health services,

236 students made suggestions such as providing access to free counselling across the country

237 and making the process of seeking help more clear, as well as increasing awareness of

238 available services. Despite the largely negative views towards mental health services, more

239 than half of the present sample mentioned that they would seek professional help in order to

240 assist a friend who was feeling suicidal. In contrast, Curtis' (2010) sample thought that

241 seeking professional help was a last resort, with many preferring to seek the support of family

242 and friends primarily.

243 There were a number of important qualitative results regarding what students believed

244 to be the cause of youth suicide. The most commonly perceived cause of suicide for young

245 New Zealanders was bullying, with nearly a third of the sample mentioning this. In

246 comparison, Heled and Read (2005) had just 4.2% of their participants mention bullying as a

247 cause of youth suicide. The high percentage reported in the present study is a result echoed

248 by recent New Zealand research by Stubbing and Gibson (2018).

249 In the present study, suggestions from participants to reduce New Zealand's youth

250 suicide rate included tackling stigma and raising awareness about mental illness and suicide,

BELIEFS ABOUT YOUTH SUICIDE

251 although how to do this was not specified. Such actions require change at a societal level to

252 challenge the negative perceptions held about people who feel suicidal. Stigma surrounding

253 suicide involves the perception that someone who attempts or completes suicide is selfish,

254 cowardly, or attention seeking, similar to stigma towards mental illness (Sheehan,

255 Nieweglowski, & Corrigan, 2017). Government-funded campaigns such as the ongoing 'Like

256 Minds, Like Mine' campaign, have tried to reduce New Zealand's stigma and discrimination

257 against people experiencing mental illness (Health Promotion Agency, 2018). Such anti-

258 stigma campaigns have resulted in modest improvements, decreasing stigmatising attitudes

259 and increasing knowledge, which in turn has been associated with an increase in help-seeking

260 behaviour (Carpiniello & Pinna, 2017).

261 Interestingly, few participants perceived media stories about suicide to be an

262 important cause of youth suicide, with participants wanting more suicide awareness, not less.

263 This is despite the fact that empirical research suggests that media reports of suicide can

264 provoke copycat suicides (Stack, 2005) and suicide contagion (Gould, Jamieson, & Romer,

265 2003). Participants in the present study seem to be endorsing the argument that public

266 awareness and education about where to seek help can help to prevent suicides (Gluckman,

267 2017), much like participants in Heled and Read's (2005) study, who suggested increasing

268 media stories about suicide could be beneficial. New Zealand's guidelines for the responsible

269 reporting of suicides encourage reports about suicide prevention (Ministry of Health, 2011).

270 Perhaps reports with this angle should be more prominent in the media than they are at

271 present, to increase public awareness of suicide and ways to get help.

272 Notably missing in participants’ descriptions of the causes of youth suicide were

273 attributions to genetic and biological causes, with participants instead emphasising situational

274 causes of suicide such as bullying and social pressures. This is consistent with Heled and

275 Read's (2005) findings. In some ways, the participants’ sole emphasis on situational factors

BELIEFS ABOUT YOUTH SUICIDE

276 is at odds with the research literature: There is, for example, strong evidence that suicidal

277 behaviour is partially heritable (Baldessarini & Hennen, 2004; Brent & Mann, 2005; Voracek

278 & Loibl, 2007). Whether this difference between the beliefs of youth and the research

279 evidence is a problem is an open question, although health professionals should certainly be

280 aware that family history of suicide is a risk factor for suicide in young people (see Agerbo,

281 Nordentoft, & Mortensen, 2002).

282 Practical Implications and Recommendations

283 Despite the finding that many students would seek help from a professional service

284 for a suicidal friend, they largely perceived such services to be inadequate for several reasons,

285 as previously mentioned. To challenge and change this perception, mental health services

286 should consider taking steps to increase accessibility. Even the perception of barriers to

287 accessibility could be problematic, because if youth believe these services are difficult to

288 access they likely will not try to access them, regardless of whether those barriers actually

289 exist or not.

290 Limitations

291 The main limitations of this study relate to the size and demographic characteristics of

292 the sample. As participation was limited to students at Massey University, just one of the

293 eight universities in New Zealand, it cannot be assumed the results of this study generalise to

294 other New Zealand university student populations or to the general population of youth

295 worldwide. The under-representation of males in the sample was problematic given that

296 males are typically over-represented in suicide statistics (Hawton, 2000); and males may hold

297 different views towards suicide and help-seeking compared to females. Such views will not

298 have been accurately encapsulated in the results of this study. Subgroup analyses comparing

299 male and female views could be conducted with the data obtained by the present study.

300 However, no analyses by gender have been presented in this report, due to the small number

BELIEFS ABOUT YOUTH SUICIDE

301 of male participants, and the fact that such analyses were not preregistered. The quantitative

302 data is openly accessible should others wish to explore these possibilities.

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

BELIEFS ABOUT YOUTH SUICIDE

326 References

327 Agerbo, E., Nordentoft, M., & Mortensen, P. B. (2002). Familial, psychiatric, and 328 socioeconomic risk factors for suicide in young people: Nested case-control study. 329 British Medical Journal, 325(7355), 74. doi:https://doi.org/10.1136/bmj.325.7355.74

330 Baldessarini, R. J., & Hennen, J. (2004). Genetics of suicide: An overview. Harvard Review 331 of Psychiatry, 12(1), 1-13. doi:https://doi.org/10.1080/10673220490425915

332 Beautrais, A. L. (2003). Suicide in New Zealand II: A review of risk factors and prevention. 333 New Zealand Medical Journal, 116(1175).

334 Brent, D. A., & Mann, J. J. (2005). Family genetic studies, suicide, and suicidal behavior. 335 Paper presented at the American Journal of Medical Genetics Part C: Seminars in 336 Medical Genetics.

337 Carpiniello, B., & Pinna, F. (2017). The reciprocal relationship between suicidality and 338 stigma. Frontiers in Psychiatry, 8(35), 1-9. doi:https://doi.org/10.3389/fpsyt.2017.00035

339 Coronial Services of New Zealand. (2018). Annual suicide statistics since 2011. Retrieved 340 from https://coronialservices.justice.govt.nz/suicide/annual-suicide-statistics-since-2011/

341 Corrigan, P. W., Angell, B. A., Davidson, L., Marcus, S. C., Salzer, M. S., Kottsieper, P., . . . 342 Stanhope, V. (2012). From adherence to self-determination: Evolution of a treatment 343 paradigm for people with serious mental illnesses. Psychiatric Services, 63(2), 169-173. 344 doi:https://doi.org/10.1176/appi.ps.201100065

345 Curtis, C. (2010). Youth perceptions of suicide and help-seeking: ‘They'd think I was weak or 346 “mental”’. Journal of Youth Studies, 13(6), 699-715. 347 doi:https://doi.org/10.1080/13676261003801747

348 Forman, J., & Damschroder, L. (2007). Qualitative content analysis. In L. Jacoby & L. A. 349 Siminoff (Eds.), Empirical methods for bioethics: A primer (pp. 39-62): Emerald Group 350 Publishing Limited.

BELIEFS ABOUT YOUTH SUICIDE

351 Gluckman, P. (2017). Youth suicide in New Zealand: A discussion paper. Retrieved from 352 http://www.pmcsa.org.nz/wp-content/uploads/17-07-26-Youth-suicide-in-New-Zealand- 353 a-Discussion-Paper.pdf

354 Gould, M., Jamieson, P., & Romer, D. (2003). Media contagion and suicide among the 355 young. American Behavioral Scientist, 46(9), 1269-1284. 356 doi:https://doi.org/10.1177%2F0002764202250670

357 Hawton, K. (2000). Sex and suicide: Gender differences in suicidal behaviour. The British 358 Journal of Psychiatry, 177(6), 484-485. doi:https://doi.org/10.1192/bjp.177.6.484

359 Head, M. L., Holman, L., Lanfear, R., Kahn, A. T., & Jennions, M. D. (2015). The extent and 360 consequences of p-hacking in science. PLoS Biology, 13(3), e1002106. 361 doi:https://doi.org/10.1371/journal.pbio.1002106

362 Health Promotion Agency. (2018). Like minds, like mine. Retrieved from 363 https://www.likeminds.org.nz/

364 Heled, E., & Read, J. (2005). Young peoples' opinions about the causes of, and solutions to, 365 New Zealand's high youth suicide rate. Suicide and Life-Threatening Behavior, 35(2), 366 170-180. doi:https://doi.org/10.1521/suli.35.2.170.62881

367 Mental Health Commission. (2012). Blueprint II: How things need to be. Retrieved from 368 https://www.hdc.org.nz/media/1075/blueprint-ii-how-things-need-to-be.pdf

369 Ministry of Health. (2011). Reporting suicide: A resource for the media. Retrieved from 370 https://www.health.govt.nz/system/files/documents/publications/reporting-suicide-a- 371 resource-for-media-dec2011.pdf

372 Ministry of Health. (2017). Indicator: Psychological distress (high or very high probability of 373 anxiety of depressive disorder, K10 score >12). Retrieved from 374 https://minhealthnz.shinyapps.io/nz-health-survey-2016-17-annual-data- 375 explorer/_w_b7c105f9/#!/explore-indicators

BELIEFS ABOUT YOUTH SUICIDE

376 Ministry of Social Development. (2016). The social report 2016. Wellington, New Zealand. 377 Retrieved from http://socialreport.msd.govt.nz/documents/2016/msd-the-social-report- 378 2016.pdf

379 New Zealand Union of Students' Associations. (2018). University counselling statistics. 380 Retrieved from 381 https://d3n8a8pro7vhmx.cloudfront.net/students/pages/295/attachments/original/154092 382 5981/_University_Counselling_Statistics.pdf?1540925981

383 Nosek, B. A., Ebersole, C. R., DeHaven, A. C., & Mellor, D. T. (2018). The preregistration 384 revolution. Proceedings of the National Academy of Sciences, 115(11), 2600-2606. 385 doi:https://doi.org/10.1073/pnas.1708274114

386 Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. 387 349(6251), aac4716. doi:https://doi.org/10.1126/science.aac4716

388 Pashler, H., & Wagenmakers, E. J. (2012). Editors’ introduction to the special section on 389 replicability in psychological science a crisis of confidence? Perspectives on 390 Psychological Science, 7(6), 528-530. 391 doi:https://doi.org/10.1177%2F1745691612465253

392 R Core Team. (2018). R: A language and environment for statistical computing. Vienna, 393 Austria: R Foundation for Statistical Computing. Retrieved from https://www.R- 394 project.org/

395 Sheehan, L., Nieweglowski, K., & Corrigan, P. W. (2017). Structures and types of stigma. In 396 W. Gaebel, W. Rössler, & N. Sartorius (Eds.), The stigma of mental illness - End of the 397 story? (pp. 43-66). Switzerland: Springer International Publishing.

398 Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: 399 Undisclosed flexibility in data collection and analysis allows presenting anything as 400 significant. Psychological Science, 22(11), 1359-1366. 401 doi:https://doi.org/10.1177%2F0956797611417632

BELIEFS ABOUT YOUTH SUICIDE

402 Stack, S. (2005). Suicide in the media: A quantitative review of studies based on nonfictional 403 stories. Suicide and Life-Threatening Behavior, 35(2), 121-133. 404 doi:https://doi.org/10.1521/suli.35.2.121.62877

405 Stallman, H. M. (2010). Psychological distress in university students: A comparison with 406 general population data. Australian Psychologist, 45(4), 249-257. 407 doi:https://doi.org/10.1080/00050067.2010.482109

408 Stubbing, J., & Gibson, K. (2018). Young people's explanations for youth suicide in New 409 Zealand: A thematic analysis. Journal of Youth Studies, 1-13. 410 doi:https://doi.org/10.1080/13676261.2018.1516862

411 Voracek, M., & Loibl, L. M. (2007). Genetics of suicide: A systematic review of twin 412 studies. Wiener klinische Wochenschrift, 119(15), 463-475. 413 doi:https://doi.org/10.1007/s00508-007-0823-2

414 World Health Organization. (2018). Suicide. Retrieved from https://www.who.int/news- 415 room/fact-sheets/detail/suicide

416

417

418

419

420

421

422

423

424

425

426

427

BELIEFS ABOUT YOUTH SUICIDE

428 Table 1

429 Participant Demographics

Variable %

Gender

Female 89

Male 10

Gender diverse 1

Ethnicity

New Zealand European/ Pākehā 76

Māori 8

Pacific Island 4

Asian 4

Other 8

Area of study

Psychology 43

Social sciences and humanities (excluding psychology) 23

Business 12

Health sciences (excluding psychology) 5

Physical or mathematical sciences 4

Other 13

430 Note. As there are 100 participants, the percentages and frequencies of participants falling in

431 each category are the same.

432